Skip to main content
ARS Home » Southeast Area » Mississippi State, Mississippi » Crop Science Research Laboratory » Genetics and Sustainable Agriculture Research » Research » Publications at this Location » Publication #261338

Title: A categorical, improper probability method for combining NDVI and LiDAR elevation information for potential cotton precision agricultural applications

item Willers, Jeffrey
item WU, JIXIANG - South Dakota State University
item Jenkins, Johnie

Submitted to: Computers and Electronics in Agriculture
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 11/27/2011
Publication Date: 3/1/2012
Citation: Willers, J.L., Wu, J., O'Hara, C., Jenkins, J.N. 2012. A categorical, improper probability method for combining NDVI and LiDAR elevation information for potential cotton precision agricultural applications. Computers and Electronics in Agriculture. 82:15-22.

Interpretive Summary: An image processing procedure was developed to combine vegetative indices derived from multispectral imagery with Light Detection and Ranging (LiDAR) derived digital elevation information collected over commercial cotton fields. The purpose of this procedure is to assist producers and field consultants. Specifically, instead of having to carry to the field at the same time, one map for crop vigor as determined by the vegetative index product and another map depicting the elevational relief, they can use procedure to make and carry only one map. From that single map, they can quickly understand how the two original input products interact to describe crop conditions at the time of the most recent multi-spectral acquisition. The combined map product is expected to be useful for improved field scouting for cotton flower bud, flower, or cotton boll damaging insect pests.

Technical Abstract: An algorithm is presented to fuse the Normalized Difference Vegetation Index (NDVI) with Light Detection and Ranging (LiDAR) elevation data to produce a map potentially useful for the site-specific scouting and pest management of several insect pests. In cotton, these pests include the Tarnished Plant Bug (Lygus lineolaris (Palisot de Beauvois) Heteroptera: Miridae) and various species of the stink bug complex (Heteroptera: Pentatomidae). The algorithm is derived from a bi-variate Gaussian density distribution, modified by several algebraic changes and applications of categorical variables. The inputs are raster representations of the NDVI and a LiDAR derived digital elevation (m) model. The development cycle leading to current implementation of the algorithm is described. During 2004, 2008 and 2009, different types of fused maps were developed. Based upon field usage while sampling for tarnished plant bugs and stink bugs over several seasons, iterative refinements were made. Since crop phenology (captured by the NDVI component) is strongly affected by water availability, which correlates with the slope and relative elevations within the field (captured by the LiDAR component), this fusion procedure produces a map that is pictorially descriptive of the relationship between cotton vigor and elevational relief in a spatial and temporal way. Since many insects’ exhibit behavioral and sensorial traits that respond to the variability of their environments, this fused map is potentially useful for geographical assessments of several cotton pests.